Confronting a solar irradiance reconstruction with solar and stellar data

Context. A recent paper by Shapiro and colleagues (2011, A&A, 529, A67) reconstructs spectral and total irradiance variations of the Sun during the holocene. Aims. In this note, we comment on why their methodology leads to large (0.5%) variations in the solar TSI on century-long time scales, in stark contrast to other reconstructions which have ≈0.1% variations. Methods. We examine the amplitude of the irradiance variations from the point of view of both solar and stellar data. Results. Shapiro et al.'s large amplitudes arise from differences between the irradiances computed from models A and C of Fontenla and colleagues, and from their explicit assumption that the radiances of the quiet Sun vary with the cosmic ray modulation potential. We suggest that the upper photosphere, as given by model A, is too cool, and discuss relative contributions of local vs. global dynamos to the magnetism and irradiance of the quiet Sun. We compare the slow (>22 yr) components of the irradiance reconstructions with secular changes in stellar photometric data that span 20 years or less, and find that the Sun, if varying with such large amplitudes, would still lie within the distribution of stellar photometric variations measured over a 10-20 year period. However, the stellar time series are individually too short to see if the reconstructed variations will remain consistent with stellar variations when observed for several decades more. Conclusions. By adopting model A, Shapiro et al. have over-estimated quiet-Sun irradiance variations by about a factor of two, based upon a re-analysis of sub-mm data from the James Clerk Maxwell telescope. But both estimates are within bounds set by current stellar data. It is therefore vital to continue accurate photometry of solar-like stars for at least another decade, to reveal secular and cyclic variations on multi-decadal time scales of direct interest to the Sun.